Fractional by Swarm

Everyone says AI will 10x your dev team. But what does the data actually say?In this episode of Fractional, Simon Obstbaum—Stanford researcher and former CTO of Crunchyroll—walks us through one of the most grounded conversations on developer productivity with AI you’ll hear.Simon’s team at Stanford built a model trained on expert code reviews to study the actual impact of AI coding tools. What they found: a 20% productivity uplift at best—with wide variance depending on language, codebase complexity, and context.We discuss:• What productivity really means in software engineering• Why legacy codebases often get worse with AI• How LLMs quietly replace pair programming for junior devs• What makes engineering teams more effective in reality—not theory• Lessons from building Crunchyroll’s global tech orgThis isn’t an AI hypefest. It’s a reality check.🧪 Participate in Stanford’s research👤 Connect with Simon🐝 Learn how Swarm helps enterprises adopt AI🎙 Fractional by Swarm

Everyone’s shipping copilots.Ragde Falcis built something else: a real agentic system—used daily by thousands of customers.In this episode of Fractional, we break down how ChatGenie:→ Hit 90%+ accuracy without model fine-tuning→ Uses multi-agent orchestration with reasoning and memory→ Built eval loops and guardrails that actually scale→ Tracks token-level usage to make the business work→ Retrofits into Meta’s stack, CRMs, and internal HR workflows→ Thinks beyond PH—and why that mindset shift mattersThis isn’t another AI demo.It’s the playbook for what actually works in production.🎧 Listen if you're serious about:Shipping agent-based systemsScaling GenAI workflowsOr building infrastructure that survives beyond the hypeBuilt something real? You’ll need real help to scale it.Swarm is a collective of AI-native consultants—builders, operators, and strategists from MAANG, YC, and Fortune 500s.If you’re building your AI roadmap and need fractional support that actually ships, let’s talk.🔗 https://swarm.work/?utm_source=fractional&utm_medium=podcast&utm_campaign=ragde_episode

"Most OKRs Are Just Tracking Failure"OKRs are supposed to drive success. But here’s what actually happens instead:1. Teams track effort, not results.2. Leaders set goals based on gut feeling.3. When things go wrong, strategy isn’t questioned—the team is."If your OKRs aren’t tied to behavior change, they’re just noise."Jeff Gothelf and I break down:1. How OKRs expose bad leadership assumptions2. Why most OKRs don’t actually measure success3. How to use OKRs to push back—without career risk4. Why customer behavior, not feature launches, should define success5. How product discovery should be continuous, not a one-time study6. Why designers, PMs, and AI teams need OKRs that track learning, not just output7. How to create a culture of continuous learning—where teams challenge assumptions, test ideas, and adapt fastBecause if your OKRs aren’t helping you challenge bad decisions, they’re useless.Jeff is a a leading voice and thought leader on customer centric product development. He helps organizations build better products and executives build the cultures that build better products. He is the co-author of the award-winning book Lean UX (now in it’s 3rd edition) and the Harvard Business Review Press book Sense & Respond. In 2020 he published the critically aclaimed book, Forever Employable.Starting off as a software designer then product manager and finally entrepreneur, Jeff now works as a keynote speaker, trainer, coach, and board advisor helping companies put the customer in the century of every decision.Most recently Jeff co-authored Who Does What By How Much?, his 5th book and 3rd with co-author Josh Seiden. This new book is the how-to guide for implementing Objectives and Key Results (OKRs) in every organization.Jeff's latest bookGet access to our network of expert AI consultancies

AI without strategy is just hype—Wardley Maps fix that.Governments, businesses, and enterprises follow AI trends without mapping the landscape, leaving them blind to risks and opportunities.That's why I sat down with Simon Wardley, the inventor of Wardley Maps, and one of the UK’s top 50 most influential people in tech, to break down why most AI strategies fail—and how mapping can give you a real advantage.In this episode of Fractional:🚀 Why AI without a map is like playing chess blindfolded📍 How Wardley Maps expose hidden risks and opportunities in AI adoption💡 The power struggles shaping AI—and why open-source AI is critical⚠️ Why governments must fight for open AI instead of corporate control🛠️ How AI tools, language, and reasoning are evolving—and what that means for strategyIf you're responsible for AI adoption in your company, this conversation will change how you think about strategy.🔹 Who is Simon Wardley?A former CEO, startup advisor (all acquired by US tech giants), and fellow of Open Europe, Simon has spent his career navigating complex systems, from AI and economics to behavioral patterns and market shifts. He pioneered Wardley Mapping, a framework used by organizations worldwide to anticipate change, outmaneuver competitors, and drive real AI strategy.

Rajiv Ayyangar is the CEO at Product Hunt. He was previously co-founder / CEO of Tandem. He also studied bioengineering at Princeton and subsequently worked at Fairchild and Yahoo. In this episode of Fractional, we discuss the Product Hunt AI roadmap, what AI products are getting the top spot, Rajiv's journey from Tandem CEO to ProductHunt CEO, and why build velocity is very important. If you're working on your GenAI roadmap, strategy, and need an implementation partner, do it with Swarm!

"AI in healthcare isn’t just hard—it’s life or death." In this episode of Fractional by Swarm, we sit down with Evelyn C. Goh—PhD candidate at the National University of Singapore (NUS) and founder/CEO of Megrez—to explore the toughest challenges in healthcare AI. Evelyn is pioneering research on the integration of AI into healthcare solutions. Through her work at Megrez, an agency specializing in medical AI/software consultancy and development, she is transforming the way healthcare organizations approach innovation. Evelyn is passionate about simplifying complex processes, making technology more accessible, and empowering individuals to create impactful clinical experiences. Join us as we dive into:✅ How to adopt AI in healthcare without losing trust✅ Why healthcare AI is so complex—and how to navigate it✅ The critical importance of solving real problems, not just chasing trends Evelyn’s insights are candid, actionable, and rooted in experience. She breaks down what it takes to build AI models that deliver real impact while meeting the demands of trust, privacy, and compliance. Whether you’re a healthcare executive or curious about AI’s transformative potential, this episode is a must-listen. 🎧 Tune in now to learn how healthcare leaders can harness AI to revolutionize patient care. 🔗 Learn more about Swarm: https://swarm.work?utm_source=spotify&utm_medium=podcast&utm_campaign=healthcare_ai_episode 🔗 Discover Megrez: https://megrez.asia/?utm_source=spotify&utm_medium=podcast&utm_campaign=healthcare_ai_episode #AIinHealthcare #HealthcareLeadership #HealthcareInnovation

Tomithy plays the role of seeker and investor at ST Telemedia - a VC/PE fund that invests in telcos, data centres, growth stage tech companies in AI, Cybersecurity and Cloud IT . He actively works with target and existing portfolio companies in terms of technical diligence, business strategy and commercial evaluation via their CTO, product & engineering teams. He has a deep interest in governance and reliability of AI systems, pushing the limits of generative AI, multi-agent AI systems, evolutionary algorithms and synthetic biology. Tomithy holds a degree in Computational Biology (Hons) from NUS, Masters in AI program from SMU and is currently pursuing a part-time PhD in AI Governance at SMU. In this episode, we talk about fragmentation and consolidation in the GenAI space, using GenAI as a business strategy, anti-patterns to avoid when implementing searchable enterprise databases, investing in GenAI, and a lot more. 🔹 Spending cycles in AI 🔹 Lessons learned from investing in AI 🔹 Using tech for commercial differentiation 🔹 Pitching during the innovation versus scaling phase 🔹 Opinions on GenAI maps 🔹 Does AI really translate to monetization? 🔹 The overpromise of enterprise databases 🔹 GenAI as the evolution of the business and as a strategy 🔹 Don’t build generic chatbots 🔹 The shortening of the tech adoption cycle 🔹 Platformatizing AI Talk to Tomithy Produced by Swarm — Hire solution architects from Fortune 100s for your AI projects.

Anand Krishnan is the Founder and Managing Partner of thinkbridge, a digital technology and business consulting firm that specializes in helping clients build and deploy tech-forward business strategies that drive outsized value creation. He is also the co-founder of Sensorscall, a senior care platform that helps seniors age-in-place. He was a former CTO of IBM and is one of the leading hives on Swarm. Anand helps growth-stage companies scale and increase their enterprise value, through the meaningful use of technology. In this episode, we talk about designing a hybrid organization for CTOs and external software engineering partners, using a transparent and outcome-first business model, inconsequential versus meaningful technology, finding your impact as a veteran consultant, and a lot more: 🔹 Scaling hybrid teams beyond “big tech” tactics 🔹 Why Thinkbridge moved from hourly to monthly pricing 🔹 Addressing the mid-market transformation gap 🔹 Avoiding “vanity tech”—what really drives impact 🔹 Redefining retention with transparency and accountability 🔹 Anand’s roadmap to agile scaling without trade-offs 🔹 Making tech teams resilient with hybrid models 🔹 Discovery phases as keys to high-value outcomes 🔹 Progress over perfection—the new rule for growth-stage orgs 🔹 AI’s role in leveling the playing field for emerging tech 🔹 Why “boring” tech like EMRs matter 🔹 Aligning tech impact with revenue for real outcomes 🔹 Purpose-driven tech—Anand’s approach to meaningful success Anand's LinkedIn:   / anandhkrishnan   Work with Anand and thinkbridge on Swarm

Dr Bernard Leong is currently the founder and CEO of Dorje AI, an enterprise AI startup focusing on building the next generation business operating system that enables companies to be productive, efficient, and adaptive. He held regional business roles as the Head of Artificial Intelligence & Machine Learning in Amazon Web Services for ASEAN, Vice President of Airbus Aerial specifically on satellite and drone services in the Asia Pacific, and senior management roles in Singapore Post as the Head of Digital Services and Group Chief Information & Digital Office in Woh Hup. Bernard is the founder and host of the Analyse Asia podcast, and is an active angel investor in web3 and AI startups. He holds a BSc in Science with Merit in National University of Singapore, and PhD in Physics from Cambridge University and has developed machine learning algorithms to search for mRNA targets in the Wellcome Trust Sanger Institute from 2003-2005. In this episode, we talk about leading the AI/ML business in AWS, AI GTM for AI consultants, creating a financial ERP startup, why change management is critical to land the biggest enterprises, and a lot more: https://www.linkedin.com/in/bleongcw/ https://www.bernardleong.com/ (0:00) - Trailer(0:30) - Bernard’s Background(2:31) - ERPless world with AI(2:55) - Human Genome Project and mRNA targeting(4:15) - Startup Experience(5:00) - Singapore Post(6:50) - Working with Apple(8:38) - Airbus Aerial(8:48) - AWS(9:20) - Digitally transforming companies(9:52) - Most enterprises use ERPs(10:18) - Why people hate ERPs(10:50) - Data silos in SAP(12:20) - Building an AI first company(13:31) - Enterprise data readiness(15:04) - Structuring data for machine learning(15:21) - Finding patterns(15:37) - Everyone loves transformation no one wants to change(16:03) - Board and C-level conversations(16:33) - Helping the biggest SEA companies(17:16) - McKinsey vs Bernard on data readiness(17:58) - What enterprises are missing out on(19:23) - Expounding on data readiness(20:45) - Finding your AI use case(21:11) - Augmenting productivity(21:20) - Increasing revenue(22:15) - Netflix and AWS machine learning(22:38) - Layers of AI implementation(24:30) - Productivity gains from modern AI(24:41) - Hyper parameter finetuning(25:20) - Dorje.AI AI powered ledger(26:55) - Financial processes into vector embeddings(28:00) - Not underestimating SAP and Oracle(28:40) - Challenging the ledger with AI and modern architecture(29:33) - Why is it hard to use SAP for businesses?(30:09) - Bernard’s ideal world with Dorje.ai(31:29) - Orchestrating data(31:48) - SaaS model upheaval(33:37) - AI GTM and Product(33:59) - Finding PMF(35:28) - Creating hockey stick growth through AI(36:24) - SAP reworks(37:03) - AI OCR built by Bernard(37:53) - Recreating the app with ChatGPT(38:26) - Refactoring the code easily(38:51) - AI coding(39:46) - Cap tables with AI(40:58) - Productivity gains with ChatGPT(41:37) - AI first mentality(43:05) - Automating things with LLMs(44:32) - AI pre-empting problems of ERPs(45:18) - One click transfer(45:47) - What Dorje means(47:46) - Competition and execution(48:42) - Conflict of Visions(49:14) - Constrained vs Unconstrained Founder Visions(50:11) - Best founders(51:07) - Reading the only book on SAP history(52:12) - Building good accounting software(52:27) - How prepared is Southeast Asia?(54:03) - What’s making the money in AI?(54:20) - Sequoia, AI Billion Hole(55:11) - Local data strategy and startup survival(56:14) - Business owner BPO in PH(57:45) - AI Consulting Advice(58:38) - Buy in and productization(59:05) - Business development in Amazon(1:00:04) - Technical POC with 50 customers(1:00:42) - Automating yourself away(1:01:32) - Sales levels conversations(1:02:09) - Creating different sales motions per AI business(1:02:49) - Iterating the product via GTM(1:03:49) - War stories in Amazon

Dominic Ligot is a data analyst, researcher, software developer, entrepreneur and technologist. He’s the Founder of CirroLytix, a social impact AI company, and Data Ethics PH, an online community focused on social issues such as data privacy, data security, AI-driven discrimination and more. He represents the Philippines as Member of the Expert Advisory Panel of the International Scientific Report on Advanced AI Safety convened by the UK Department of Science, Innovation, and Technology. He led the development of the nationwide data science education program Project SPARTA. He also co-founded the Analytics and AI Association of the Philippines (AAP). Dominic is a three-time global winner of the NASA and ESA International Space Apps Challenges.  In this episode, we talk about Gen AI adoption in the Philippines, using AI in the humanitarian sector, go-to-market strategies for AI consultants, and AI safety. Dominic's website Dominic's LinkedIn Ex-MAANG, YC AI Implementation teams to get your AI journey unblocked